Statistical Modelling of COVID-19 Outbreak in Italy

05 May 2020



Nonlinear growth models

Nonlinear growth models represent an instance of nonlinear regression models, a class of models taking the general form \[ y = \mu(x, \theta) + \epsilon, \] where \(\mu(x, \theta)\) is the mean function which depends on a possibly vector-valued parameter \(\theta\), and a possibly vector-valued predictor \(x\). The stochastic component \(\epsilon\) represents the error with mean zero and constant variance. Usually, a Gaussian distribution is also assumed for the error term.

By defining the mean function \(\mu(x, \theta)\) we may obtain several different models, all characterized by the fact that parameters \(\theta\) enter in a nonlinear way into the equation. Parameters are usually estimated by nonlinear least squares which aims at minimizing the residual sum of squares.

Exponential

\[ \mu(x) = \theta_1 \exp\{\theta_2 x\} \] where \(\theta_1\) is the value at the origin (i.e. \(\mu(x=0)\)), and \(\theta_2\) represents the (constant) relative ratio of change (i.e. \(\frac{d\mu(x)}{dx }\frac{1}{\mu(x)} = \theta_2\)). Thus, the model describes an increasing (exponential growth if \(\theta_2 > 0\)) or decreasing (exponential decay if \(\theta_2 < 0\)) trend with constant relative rate.

Logistic

\[ \mu(x) = \frac{\theta_1}{1+\exp\{(\theta_2 - x)/\theta_3\}} \] where \(\theta_1\) is the upper horizontal asymptote, \(\theta_2\) represents the x-value at the inflection point of the symmetric growth curve, and \(\theta_3\) represents a scale parameter (and \(1/\theta_3\) is the growth-rate parameter that controls how quickly the curve approaches the upper asymptote).

Gompertz

\[ \mu(x) = \theta_1 \exp\{-\theta_2 \theta_3^x\} \] where \(\theta_1\) is the horizontal asymptote, \(\theta_2\) represents the value of the function at \(x = 0\) (displacement along the x-axis), and \(\theta_3\) represents a scale parameter.

The difference between the logistic and Gompertz functions is that the latter is not symmetric around the inflection point.

Richards

\[ \mu(x) = \theta_1 (1 - \exp\{-\theta_2 x\})^{\theta_3} \] where \(\theta_1\) is the horizontal asymptote, \(\theta_2\) represents the rate of growth, and \(\theta_3\) in part determines the point of inflection on the y-axis.

Data

Dipartimento della Protezione Civile: COVID-19 Italia - Monitoraggio della situazione http://arcg.is/C1unv

Source: https://github.com/pcm-dpc/COVID-19

## # Dati COVID-19 Italia
## 
## ## Avvisi
## 
## ```diff
## - 02/05/2020: dati Regione Lombardia ricalcolati 329 decessi (47 di oggi e 282 da riconteggio di aprile)
## - 01/05/2020: dati Regione Lazio ricalcolati 41 decessi (8 nelle ultime 48 ore e 33 ad aprile)
## - 26/04/2020: dati Regione Valle d'Aosta ricalcolati (casi testati)
## - 24/04/2020: dati Regione Sardegna ricalcolati (1.237 tamponi aggiunti)
## - 24/04/2020: dati Regione Friuli Venezia Giulia in fase di revisione su dimessi/guariti
## - 23/04/2020: dati Regione Lazio parziali (casi testati non completi)
## - 23/04/2020: dati Regione Campania parziali (casi testati non aggiornati)
## - 21/04/2020: dati Regione Lombardia parziali (casi testati non aggiornati)
## - 20/04/2020: dati Regione Lombardia ricalcolati (ricalcolo di casi testati - eliminazione duplicati)
## - 15/04/2020: dati Regione Friuli Venezia Giulia ricalcolati (ricalcolo di isolamento domiciliare e dimessi/guariti)
## - 12/04/2020: dati P.A. Bolzano ricalcolati (ricalcolo dati guariti -110 rispetto a ieri)
## - 10/04/2020: dati Regione Molise parziali (dato tamponi non aggiornato)
## - 29/03/2020: dati Regione Emilia-Romagna parziali (dato tamponi non aggiornato)
## - 26/03/2020: dati Regione Piemonte parziali (-50 deceduti - comunicazione tardiva)
## - 18/03/2020: dati Regione Campania non pervenuti
## - 18/03/2020: dati Provincia di Parma non pervenuti
## - 17/03/2020: dati Provincia di Rimini non aggiornati
## - 16/03/2020: dati P.A. Trento e Puglia non pervenuti
## - 11/03/2020: dati Regione Abruzzo non pervenuti
## - 10/03/2020: dati Regione Lombardia parziali
## - 07/03/2020: dati Brescia +300 esiti positivi
## ```
url = "https://raw.githubusercontent.com/pcm-dpc/COVID-19/master/dati-andamento-nazionale/dpc-covid19-ita-andamento-nazionale.csv"
COVID19 <- read.csv(file = url, stringsAsFactors = FALSE)
COVID19$data <- as.Date(COVID19$data)
# DT::datatable(COVID19)


Modelling total infected

# create data for analysis
data = data.frame(date = COVID19$data,
                  y = COVID19$totale_casi,
                                    dy = reldiff(COVID19$totale_casi))
data$x = as.numeric(data$date) - min(as.numeric(data$date)) + 1
DT::datatable(data, options = list("pageLength" = 5))

Estimation

Exponential

mod1_start = lm(log(y) ~ x, data = data)
b = unname(coef(mod1_start))
start = list(th1 = exp(b[1]), th2 = b[2])
mod1 = nls(y ~ exponential(x, th1, th2), data = data, start = start)
summary(mod1)
## 
## Formula: y ~ exponential(x, th1, th2)
## 
## Parameters:
##         Estimate   Std. Error t value           Pr(>|t|)    
## th1 26470.147045  2781.711616   9.516 0.0000000000000302 ***
## th2     0.031768     0.001776  17.884            < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 25800 on 70 degrees of freedom
## 
## Number of iterations to convergence: 12 
## Achieved convergence tolerance: 0.000006982

Logistic

mod2 = nls(y ~ SSlogis(x, Asym, xmid, scal), data = data)
summary(mod2)
## 
## Formula: y ~ SSlogis(x, Asym, xmid, scal)
## 
## Parameters:
##         Estimate  Std. Error t value Pr(>|t|)    
## Asym 209849.3245   2249.5266   93.29   <2e-16 ***
## xmid     37.8496      0.3483  108.66   <2e-16 ***
## scal      9.2610      0.2600   35.62   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5618 on 69 degrees of freedom
## 
## Number of iterations to convergence: 0 
## Achieved convergence tolerance: 0.0000009482

Gompertz

mod3 = nls(y ~ SSgompertz(x, Asym, b2, b3), data = data)
# start = list(Asym = coef(mod2)[1])
# tmp = list(y = log(log(start$Asym) - log(data$y)), x = data$x)
# b = unname(coef(lm(y ~ x, data = tmp)))
# start = c(start, c(b2 = exp(b[1]), b3 = exp(b[2])))
# mod3 = nls(y ~ SSgompertz(x, Asym, b2, b3), data = data, start = start,
#            control = nls.control(maxiter = 1000))
summary(mod3)
## 
## Formula: y ~ SSgompertz(x, Asym, b2, b3)
## 
## Parameters:
##           Estimate    Std. Error t value Pr(>|t|)    
## Asym 231511.633541   1385.507247  167.09   <2e-16 ***
## b2        7.878834      0.165231   47.68   <2e-16 ***
## b3        0.940156      0.000735 1279.21   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1952 on 69 degrees of freedom
## 
## Number of iterations to convergence: 0 
## Achieved convergence tolerance: 0.0000002013

Richards

richards <- function(x, th1, th2, th3) th1*(1 - exp(-th2*x))^th3
Loss  <- function(th, y, x) sum((y - richards(x, th[1], th[2], th[3]))^2) 
start <- optim(par = c(coef(mod2)[1], 0.001, 1), fn = Loss, 
               y = data$y, x = data$x)$par
names(start) <- c("th1", "th2", "th3")
mod4 = nls(y ~ richards(x, th1, th2, th3), data = data, start = start,
           # trace = TRUE, algorithm = "plinear", 
           control = nls.control(maxiter = 1000, tol = 0.1))
# algorithm is not converging... 
summary(mod4)
## 
## Formula: y ~ richards(x, th1, th2, th3)
## 
## Parameters:
##           Estimate     Std. Error t value Pr(>|t|)    
## th1 239559.4004823   1430.1921559  167.50   <2e-16 ***
## th2      0.0535338      0.0007407   72.28   <2e-16 ***
## th3      5.6070619      0.1228977   45.62   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1505 on 69 degrees of freedom
## 
## Number of iterations to convergence: 2 
## Achieved convergence tolerance: 0.002399
# library(nlmrt)
# mod4 = nlxb(y ~ th1*(1 - exp(-th2*x))^th3, 
#             data = data, start = start, trace = TRUE)

Models comparison

models = list("Exponential model" = mod1, 
              "Logistic model" = mod2, 
              "Gompertz model" = mod3,
              "Richards model" = mod4)
tab = data.frame(loglik = sapply(models, logLik),
                 df = sapply(models, function(m) attr(logLik(m), "df")),
                 Rsquare = sapply(models, function(m) 
                                  cor(data$y, fitted(m))^2),
                 AIC = sapply(models, AIC),
                 AICc = sapply(models, AICc),
                 BIC = sapply(models, BIC))
sel <- apply(tab[,4:6], 2, which.min)
tab$"" <- sapply(tabulate(sel, nbins = length(models))+1, symnum,
                 cutpoints = 0:4, symbols = c("", "*", "**", "***"))
knitr::kable(tab)
loglik df Rsquare AIC AICc BIC
Exponential model -832.5317 3 0.8992689 1671.063 1671.416 1677.893
Logistic model -722.2642 4 0.9955528 1452.528 1453.125 1461.635
Gompertz model -646.1523 4 0.9993836 1300.305 1300.902 1309.411
Richards model -627.4071 4 0.9996314 1262.814 1263.411 1271.921 ***
ggplot(data, aes(x = date, y = y)) + 
  geom_point() +
  geom_line(aes(y = fitted(mod1), color = "Exponential")) +
  geom_line(aes(y = fitted(mod2), color = "Logistic")) +
  geom_line(aes(y = fitted(mod3), color = "Gompertz")) +
  geom_line(aes(y = fitted(mod4), color = "Richards")) +
  labs(x = "", y = "Infected", color = "Model") +
  scale_color_manual(values = cols) +
  scale_y_continuous(breaks = seq(0, coef(mod2)[1], by = 10000),
                     minor_breaks = seq(0, coef(mod2)[1], by = 5000)) +
  scale_x_date(date_breaks = "2 day", date_labels =  "%b%d",
               minor_breaks = "1 day") +
  theme_bw() +
  theme(legend.position = "top",
        axis.text.x = element_text(angle=60, hjust=1))

last_plot() +
  scale_y_continuous(trans = "log10", limits = c(100,NA)) +
  labs(y = "Infected (log10 scale)")

Predictions

Point estimates

df = data.frame(x = seq(min(data$x), max(data$x)+14))
df = cbind(df, date = as.Date(df$x, origin = data$date[1]-1),
               fit1 = predict(mod1, newdata = df),
               fit2 = predict(mod2, newdata = df),
               fit3 = predict(mod3, newdata = df),
               fit4 = predict(mod4, newdata = df))
ylim = c(0, max(df[,c("fit2", "fit3")]))
ggplot(data, aes(x = date, y = y)) + 
  geom_point() +
  geom_line(data = df, aes(x = date, y = fit1, color = "Exponential")) +
  geom_line(data = df, aes(x = date, y = fit2, color = "Logistic")) +
  geom_line(data = df, aes(x = date, y = fit3, color = "Gompertz")) +
  geom_line(data = df, aes(x = date, y = fit4, color = "Richards")) +
  coord_cartesian(ylim = ylim) +
  labs(x = "", y = "Infected", color = "Model") +
  scale_y_continuous(breaks = seq(0, max(ylim), by = 10000),
                     minor_breaks = seq(0, max(ylim), by = 5000)) +
  scale_x_date(date_breaks = "2 day", date_labels =  "%b%d",
               minor_breaks = "1 day") +
  scale_color_manual(values = cols) +
  theme_bw() +
  theme(legend.position = "top",
        axis.text.x = element_text(angle=60, hjust=1))

Prediction intervals

# compute prediction using Moving Block Bootstrap (MBB) for nls
df = data.frame(x = seq(min(data$x), max(data$x)+14))
df = cbind(df, date = as.Date(df$x, origin = data$date[1]-1))

pred1 = cbind(df, "fit" = predict(mod1, newdata = df))
pred1[df$x > max(data$x), c("lwr", "upr")] = predictMBB.nls(mod1, df[df$x > max(data$x),])[,2:3]

pred2 = cbind(df, "fit" = predict(mod2, newdata = df))
pred2[df$x > max(data$x), c("lwr", "upr")] = predictMBB.nls(mod2, df[df$x > max(data$x),])[,2:3]

pred3 = cbind(df, "fit" = predict(mod3, newdata = df))
pred3[df$x > max(data$x), c("lwr", "upr")] = predictMBB.nls(mod3, df[df$x > max(data$x),])[,2:3]

pred4 = cbind(df, "fit" = predict(mod4, newdata = df))
pred4[df$x > max(data$x), c("lwr", "upr")] = predictMBB.nls(mod4, df[df$x > max(data$x),])[,2:3]

# predictions for next day
pred = rbind(subset(pred1, x == max(data$x)+1, select = 2:5),
             subset(pred2, x == max(data$x)+1, select = 2:5),
             subset(pred3, x == max(data$x)+1, select = 2:5),
             subset(pred4, x == max(data$x)+1, select = 2:5))
print(pred, digits = 3)
##           date    fit    lwr    upr
## 73  2020-05-06 269108 203608 332868
## 731 2020-05-06 205237 191756 216216
## 732 2020-05-06 212199 207309 216617
## 733 2020-05-06 213804 210241 217375

ylim = c(0, max(pred2$upr, pred3$upr, na.rm=TRUE))
ggplot(data, aes(x = date, y = y)) + 
  geom_point() +
  geom_line(data = pred1, aes(x = date, y = fit, color = "Exponential")) +
  geom_line(data = pred2, aes(x = date, y = fit, color = "Logistic")) +
  geom_line(data = pred3, aes(x = date, y = fit, color = "Gompertz")) +
  geom_line(data = pred4, aes(x = date, y = fit, color = "Richards")) +
  geom_ribbon(data = pred1, aes(x = date, ymin = lwr, ymax = upr), 
              inherit.aes = FALSE, fill = cols[1], alpha=0.3) +
  geom_ribbon(data = pred2, aes(x = date, ymin = lwr, ymax = upr), 
              inherit.aes = FALSE, fill = cols[2], alpha=0.3) +
  geom_ribbon(data = pred3, aes(x = date, ymin = lwr, ymax = upr),
              inherit.aes = FALSE, fill = cols[3], alpha=0.3) +
  geom_ribbon(data = pred4, aes(x = date, ymin = lwr, ymax = upr),
              inherit.aes = FALSE, fill = cols[4], alpha=0.3) +
  coord_cartesian(ylim = c(0, max(ylim))) +
  labs(x = "", y = "Infected", color = "Model") +
  scale_y_continuous(minor_breaks = seq(0, max(ylim), by = 10000)) +
  scale_x_date(date_breaks = "2 day", date_labels =  "%b%d",
               minor_breaks = "1 day") +
  scale_color_manual(values = cols) +
  theme_bw() +
  theme(legend.position = "top",
        axis.text.x = element_text(angle=60, hjust=1))

Modelling total deceased

# create data for analysis
data = data.frame(date = COVID19$data,
                  y = COVID19$deceduti,
                                    dy = reldiff(COVID19$deceduti))
data$x = as.numeric(data$date) - min(as.numeric(data$date)) + 1
DT::datatable(data, options = list("pageLength" = 5))

Estimation

Exponential

mod1_start = lm(log(y) ~ x, data = data)
b = unname(coef(mod1_start))
start = list(th1 = exp(b[1]), th2 = b[2])
exponential <- function(x, th1, th2) th1 * exp(th2 * x)
mod1 = nls(y ~ exponential(x, th1, th2), data = data, start = start)
summary(mod1)
## 
## Formula: y ~ exponential(x, th1, th2)
## 
## Parameters:
##        Estimate  Std. Error t value         Pr(>|t|)    
## th1 2806.761801  324.617730   8.646 0.00000000000118 ***
## th2    0.035374    0.001921  18.416          < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3398 on 70 degrees of freedom
## 
## Number of iterations to convergence: 13 
## Achieved convergence tolerance: 0.000005307

Logistic

mod2 = nls(y ~ SSlogis(x, Asym, xmid, scal), data = data)
summary(mod2)
## 
## Formula: y ~ SSlogis(x, Asym, xmid, scal)
## 
## Parameters:
##        Estimate Std. Error t value Pr(>|t|)    
## Asym 28922.0948   323.9020   89.29   <2e-16 ***
## xmid    40.9134     0.3380  121.06   <2e-16 ***
## scal     8.8013     0.2462   35.76   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 740.6 on 69 degrees of freedom
## 
## Number of iterations to convergence: 0 
## Achieved convergence tolerance: 0.000002102

Gompertz

mod3 = nls(y ~ SSgompertz(x, Asym, b2, b3), data = data)
# manually set starting values
# start = list(Asym = coef(mod2)[1])
# tmp = list(y = log(log(start$Asym) - log(data$y)), x = data$x)
# b = unname(coef(lm(y ~ x, data = tmp)))
# start = c(start, c(b2 = exp(b[1]), b3 = exp(b[2])))
# mod3 = nls(y ~ SSgompertz(x, Asym, b2, b3), data = data, start = start, 
#            control = nls.control(maxiter = 10000))
summary(mod3)
## 
## Formula: y ~ SSgompertz(x, Asym, b2, b3)
## 
## Parameters:
##           Estimate    Std. Error t value Pr(>|t|)    
## Asym 32261.0839060   201.6462563   160.0   <2e-16 ***
## b2      10.5031726     0.2482884    42.3   <2e-16 ***
## b3       0.9382935     0.0007479  1254.6   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 249.1 on 69 degrees of freedom
## 
## Number of iterations to convergence: 0 
## Achieved convergence tolerance: 0.000002506

Richards

richards <- function(x, th1, th2, th3) th1*(1 - exp(-th2*x))^th3
Loss  <- function(th, y, x) sum((y - richards(x, th[1], th[2], th[3]))^2) 
start <- optim(par = c(coef(mod2)[1], 0.001, 1), fn = Loss, 
               y = data$y, x = data$x)$par
names(start) <- c("th1", "th2", "th3")
mod4 = nls(y ~ richards(x, th1, th2, th3), data = data, start = start,
           # trace = TRUE, algorithm = "port", 
           control = nls.control(maxiter = 1000))
summary(mod4)
## 
## Formula: y ~ richards(x, th1, th2, th3)
## 
## Parameters:
##         Estimate   Std. Error t value Pr(>|t|)    
## th1 33139.258501   204.245085  162.25   <2e-16 ***
## th2     0.057387     0.000764   75.11   <2e-16 ***
## th3     7.966468     0.194332   40.99   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 201.2 on 69 degrees of freedom
## 
## Number of iterations to convergence: 4 
## Achieved convergence tolerance: 0.000003586

Models comparison

models = list("Exponential model" = mod1, 
              "Logistic model" = mod2, 
              "Gompertz model" = mod3,
              "Richards model" = mod4)
tab = data.frame(loglik = sapply(models, logLik),
                 df = sapply(models, function(m) attr(logLik(m), "df")),
                 Rsquare = sapply(models, function(m) 
                                  cor(data$y, fitted(m))^2),
                 AIC = sapply(models, AIC),
                 AICc = sapply(models, AICc),
                 BIC = sapply(models, BIC))
sel <- apply(tab[,4:6], 2, which.min)
tab$"" <- sapply(tabulate(sel, nbins = length(models))+1, symnum,
                 cutpoints = 0:4, symbols = c("", "*", "**", "***"))
knitr::kable(tab)
loglik df Rsquare AIC AICc BIC
Exponential model -686.5865 3 0.9110923 1379.1730 1379.5259 1386.0030
Logistic model -576.3666 4 0.9960487 1160.7331 1161.3302 1169.8398
Gompertz model -497.9298 4 0.9994823 1003.8597 1004.4567 1012.9664
Richards model -482.5457 4 0.9996560 973.0914 973.6885 982.1981 ***
ggplot(data, aes(x = date, y = y)) + 
  geom_point() +
  geom_line(aes(y = fitted(mod1), color = "Exponential")) +
  geom_line(aes(y = fitted(mod2), color = "Logistic")) +
  geom_line(aes(y = fitted(mod3), color = "Gompertz")) +
  geom_line(aes(y = fitted(mod4), color = "Richards")) +
  labs(x = "", y = "Deceased", color = "Model") +
  scale_color_manual(values = cols) +
  scale_y_continuous(breaks = seq(0, coef(mod2)[1], by = 1000),
                     minor_breaks = seq(0, coef(mod2)[1], by = 500)) +
  scale_x_date(date_breaks = "2 day", date_labels =  "%b%d",
               minor_breaks = "1 day") +
  theme_bw() +
  theme(legend.position = "top",
        axis.text.x = element_text(angle=60, hjust=1))

last_plot() +
  scale_y_continuous(trans = "log10", limits = c(10,NA)) +
  labs(y = "Deceased (log10 scale)")

Predictions

Point estimates

df = data.frame(x = seq(min(data$x), max(data$x)+14))
df = cbind(df, date = as.Date(df$x, origin = data$date[1]-1),
               fit1 = predict(mod1, newdata = df),
               fit2 = predict(mod2, newdata = df),
               fit3 = predict(mod3, newdata = df),
               fit4 = predict(mod4, newdata = df))
ylim = c(0, max(df[,-(1:3)]))
ggplot(data, aes(x = date, y = y)) + 
  geom_point() +
  geom_line(data = df, aes(x = date, y = fit1, color = "Exponential")) +
  geom_line(data = df, aes(x = date, y = fit2, color = "Logistic")) +
  geom_line(data = df, aes(x = date, y = fit3, color = "Gompertz")) +
  geom_line(data = df, aes(x = date, y = fit4, color = "Richards")) +
  coord_cartesian(ylim = ylim) +
  labs(x = "", y = "Deceased", color = "Model") +
  scale_y_continuous(breaks = seq(0, max(ylim), by = 1000),
                     minor_breaks = seq(0, max(ylim), by = 1000)) +
  scale_x_date(date_breaks = "2 day", date_labels =  "%b%d",
               minor_breaks = "1 day") +
  scale_color_manual(values = cols) +
  theme_bw() +
  theme(legend.position = "top",
        axis.text.x = element_text(angle=60, hjust=1))

Prediction intervals

# compute prediction using Moving Block Bootstrap (MBB) for nls
df = data.frame(x = seq(min(data$x), max(data$x)+14))
df = cbind(df, date = as.Date(df$x, origin = data$date[1]-1))

pred1 = cbind(df, "fit" = predict(mod1, newdata = df))
pred1[df$x > max(data$x), c("lwr", "upr")] = predictMBB.nls(mod1, df[df$x > max(data$x),])[,2:3]

pred2 = cbind(df, "fit" = predict(mod2, newdata = df))
pred2[df$x > max(data$x), c("lwr", "upr")] = predictMBB.nls(mod2, df[df$x > max(data$x),])[,2:3]

pred3 = cbind(df, "fit" = predict(mod3, newdata = df))
pred3[df$x > max(data$x), c("lwr", "upr")] = predictMBB.nls(mod3, df[df$x > max(data$x),])[,2:3]

pred4 = cbind(df, "fit" = predict(mod4, newdata = df))
pred4[df$x > max(data$x), c("lwr", "upr")] = predictMBB.nls(mod4, df[df$x > max(data$x),])[,2:3]

# predictions for next day
pred = rbind(subset(pred1, x == max(data$x)+1, select = 2:5),
             subset(pred2, x == max(data$x)+1, select = 2:5),
             subset(pred3, x == max(data$x)+1, select = 2:5),
             subset(pred4, x == max(data$x)+1, select = 2:5))
print(pred, digits = 3)
##           date   fit   lwr   upr
## 73  2020-05-06 37127 28087 46452
## 731 2020-05-06 28186 26346 29599
## 732 2020-05-06 29177 28625 29804
## 733 2020-05-06 29342 28842 29894

ylim = c(0, max(pred2$upr, pred3$upr, na.rm=TRUE))
ggplot(data, aes(x = date, y = y)) + 
  geom_point() +
  geom_line(data = pred1, aes(x = date, y = fit, color = "Exponential")) +
  geom_line(data = pred2, aes(x = date, y = fit, color = "Logistic")) +
  geom_line(data = pred3, aes(x = date, y = fit, color = "Gompertz")) +
  geom_line(data = pred4, aes(x = date, y = fit, color = "Richards")) +
  geom_ribbon(data = pred1, aes(x = date, ymin = lwr, ymax = upr), 
              inherit.aes = FALSE, fill = cols[1], alpha=0.3) +
  geom_ribbon(data = pred2, aes(x = date, ymin = lwr, ymax = upr), 
              inherit.aes = FALSE, fill = cols[2], alpha=0.3) +
  geom_ribbon(data = pred3, aes(x = date, ymin = lwr, ymax = upr),
              inherit.aes = FALSE, fill = cols[3], alpha=0.3) +
  geom_ribbon(data = pred4, aes(x = date, ymin = lwr, ymax = upr),
              inherit.aes = FALSE, fill = cols[4], alpha=0.3) +
  coord_cartesian(ylim = c(0, max(ylim))) +
  labs(x = "", y = "Deceased", color = "Model") +
  scale_y_continuous(minor_breaks = seq(0, max(ylim), by = 1000)) +
  scale_x_date(date_breaks = "2 day", date_labels =  "%b%d",
               minor_breaks = "1 day") +
  scale_color_manual(values = cols) +
  theme_bw() +
  theme(legend.position = "top",
        axis.text.x = element_text(angle=60, hjust=1))

Modelling recovered

# create data for analysis
data = data.frame(date = COVID19$data,
                  y = COVID19$dimessi_guariti,
                                    dy = reldiff(COVID19$dimessi_guariti))
data$x = as.numeric(data$date) - min(as.numeric(data$date)) + 1
DT::datatable(data, options = list("pageLength" = 5))

Estimation

Exponential

mod1_start = lm(log(y) ~ x, data = data)
b = unname(coef(mod1_start))
start = list(th1 = exp(b[1]), th2 = b[2])
exponential <- function(x, th1, th2) th1 * exp(th2 * x)
mod1 = nls(y ~ exponential(x, th1, th2), data = data, start = start)
summary(mod1)
## 
## Formula: y ~ exponential(x, th1, th2)
## 
## Parameters:
##        Estimate  Std. Error t value Pr(>|t|)    
## th1 2592.063497  212.644148   12.19   <2e-16 ***
## th2    0.050154    0.001295   38.72   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3929 on 70 degrees of freedom
## 
## Number of iterations to convergence: 11 
## Achieved convergence tolerance: 0.000005969

Logistic

mod2 = nls(y ~ SSlogis(x, Asym, xmid, scal), data = data)
summary(mod2)
## 
## Formula: y ~ SSlogis(x, Asym, xmid, scal)
## 
## Parameters:
##         Estimate  Std. Error t value Pr(>|t|)    
## Asym 114340.4463   2957.3036   38.66   <2e-16 ***
## xmid     59.6461      0.6841   87.19   <2e-16 ***
## scal     11.7060      0.2455   47.69   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1225 on 69 degrees of freedom
## 
## Number of iterations to convergence: 0 
## Achieved convergence tolerance: 0.000002314

Gompertz

mod3 = nls(y ~ SSgompertz(x, Asym, b2, b3), data = data)
summary(mod3)
## 
## Formula: y ~ SSgompertz(x, Asym, b2, b3)
## 
## Parameters:
##            Estimate     Std. Error t value Pr(>|t|)    
## Asym 236677.7554253  10654.0802439   22.21   <2e-16 ***
## b2        8.0153899      0.1239413   64.67   <2e-16 ***
## b3        0.9715785      0.0007555 1285.97   <2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 717.1 on 69 degrees of freedom
## 
## Number of iterations to convergence: 0 
## Achieved convergence tolerance: 0.0000009936

Richards

richards <- function(x, th1, th2, th3) th1*(1 - exp(-th2*x))^th3
Loss  <- function(th, y, x) sum((y - richards(x, th[1], th[2], th[3]))^2) 
start <- optim(par = c(coef(mod2)[1], 0.001, 1), fn = Loss, 
               y = data$y, x = data$x)$par
names(start) <- c("th1", "th2", "th3")
mod4 = nls(y ~ richards(x, th1, th2, th3), data = data, start = start,
           # trace = TRUE, # algorithm = "port", 
           control = nls.control(maxiter = 1000))
summary(mod4)
## 
## Formula: y ~ richards(x, th1, th2, th3)
## 
## Parameters:
##          Estimate    Std. Error t value      Pr(>|t|)    
## th1 504948.778444  67644.479140   7.465 0.00000000019 ***
## th2      0.013938      0.001222  11.408       < 2e-16 ***
## th3      3.845936      0.143270  26.844       < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 679.4 on 69 degrees of freedom
## 
## Number of iterations to convergence: 18 
## Achieved convergence tolerance: 0.000004724

Models comparison

models = list("Exponential model" = mod1, 
              "Logistic model" = mod2, 
              "Gompertz model" = mod3,
              "Richards model" = mod4)
tab = data.frame(loglik = sapply(models, logLik),
                 df = sapply(models, function(m) attr(logLik(m), "df")),
                 Rsquare = sapply(models, function(m) 
                                  cor(data$y, fitted(m))^2),
                 AIC = sapply(models, AIC),
                 AICc = sapply(models, AICc),
                 BIC = sapply(models, BIC))
sel <- apply(tab[,4:6], 2, which.min)
tab$"" <- sapply(tabulate(sel, nbins = length(models))+1, symnum,
                 cutpoints = 0:4, symbols = c("", "*", "**", "***"))
knitr::kable(tab)
loglik df Rsquare AIC AICc BIC
Exponential model -697.0349 3 0.9831760 1400.070 1400.423 1406.900
Logistic model -612.5893 4 0.9983024 1233.179 1233.776 1242.285
Gompertz model -574.0430 4 0.9993277 1156.086 1156.683 1165.193
Richards model -570.1584 4 0.9993795 1148.317 1148.914 1157.423 ***
ggplot(data, aes(x = date, y = y)) + 
  geom_point() +
  geom_line(aes(y = fitted(mod1), color = "Exponential")) +
  geom_line(aes(y = fitted(mod2), color = "Logistic")) +
  geom_line(aes(y = fitted(mod3), color = "Gompertz")) +
  geom_line(aes(y = fitted(mod4), color = "Richards")) +
  labs(x = "", y = "Recovered", color = "Model") +
  scale_color_manual(values = cols) +
  scale_y_continuous(breaks = seq(0, coef(mod2)[1], by = 1000),
                     minor_breaks = seq(0, coef(mod2)[1], by = 500)) +
  scale_x_date(date_breaks = "2 day", date_labels =  "%b%d",
               minor_breaks = "1 day") +
  theme_bw() +
  theme(legend.position = "top",
        axis.text.x = element_text(angle=60, hjust=1))

last_plot() +
  scale_y_continuous(trans = "log10", limits = c(10,NA)) +
  labs(y = "Recovered (log10 scale)")

Predictions

Point estimates

df = data.frame(x = seq(min(data$x), max(data$x)+14))
df = cbind(df, date = as.Date(df$x, origin = data$date[1]-1),
               fit1 = predict(mod1, newdata = df),
               fit2 = predict(mod2, newdata = df),
               fit3 = predict(mod3, newdata = df),
               fit4 = predict(mod4, newdata = df))
ylim = c(0, max(df[,-(1:3)]))
ggplot(data, aes(x = date, y = y)) + 
  geom_point() + 
  geom_line(data = df, aes(x = date, y = fit1, color = "Exponential")) +
  geom_line(data = df, aes(x = date, y = fit2, color = "Logistic")) +
  geom_line(data = df, aes(x = date, y = fit3, color = "Gompertz")) +
  geom_line(data = df, aes(x = date, y = fit4, color = "Richards")) +
  coord_cartesian(ylim = ylim) +
  labs(x = "", y = "Recovered", color = "Model") +
  scale_y_continuous(breaks = seq(0, max(ylim), by = 1000),
                     minor_breaks = seq(0, max(ylim), by = 1000)) +
  scale_x_date(date_breaks = "2 day", date_labels =  "%b%d",
               minor_breaks = "1 day") +
  scale_color_manual(values = cols) +
  theme_bw() +
  theme(legend.position = "top",
        axis.text.x = element_text(angle=60, hjust=1))

Prediction intervals

# compute prediction using Moving Block Bootstrap (MBB) for nls
df = data.frame(x = seq(min(data$x), max(data$x)+14))
df = cbind(df, date = as.Date(df$x, origin = data$date[1]-1))

pred1 = cbind(df, "fit" = predict(mod1, newdata = df))
pred1[df$x > max(data$x), c("lwr", "upr")] = predictMBB.nls(mod1, df[df$x > max(data$x),])[,2:3]

pred2 = cbind(df, "fit" = predict(mod2, newdata = df))
pred2[df$x > max(data$x), c("lwr", "upr")] = predictMBB.nls(mod2, df[df$x > max(data$x),])[,2:3]

pred3 = cbind(df, "fit" = predict(mod3, newdata = df))
pred3[df$x > max(data$x), c("lwr", "upr")] = predictMBB.nls(mod3, df[df$x > max(data$x),])[,2:3]

pred4 = cbind(df, "fit" = predict(mod4, newdata = df))
pred4[df$x > max(data$x), c("lwr", "upr")] = predictMBB.nls(mod4, df[df$x > max(data$x),])[,2:3]

# predictions for next day
pred = rbind(subset(pred1, x == max(data$x)+1, select = 2:5),
             subset(pred2, x == max(data$x)+1, select = 2:5),
             subset(pred3, x == max(data$x)+1, select = 2:5),
             subset(pred4, x == max(data$x)+1, select = 2:5))
print(pred, digits = 3)
##           date    fit   lwr    upr
## 73  2020-05-06 100858 90472 111157
## 731 2020-05-06  86650 82725  89673
## 732 2020-05-06  89111 87166  90695
## 733 2020-05-06  89922 88268  91453

ylim = c(0, max(pred2$upr, pred3$upr, na.rm=TRUE))
ggplot(data, aes(x = date, y = y)) + 
  geom_point() +
  geom_line(data = pred1, aes(x = date, y = fit, color = "Exponential")) +
  geom_line(data = pred2, aes(x = date, y = fit, color = "Logistic")) +
  geom_line(data = pred3, aes(x = date, y = fit, color = "Gompertz")) +
  geom_line(data = pred4, aes(x = date, y = fit, color = "Richards")) +
  geom_ribbon(data = pred1, aes(x = date, ymin = lwr, ymax = upr), 
              inherit.aes = FALSE, fill = cols[1], alpha=0.3) +
  geom_ribbon(data = pred2, aes(x = date, ymin = lwr, ymax = upr), 
              inherit.aes = FALSE, fill = cols[2], alpha=0.3) +
  geom_ribbon(data = pred3, aes(x = date, ymin = lwr, ymax = upr),
              inherit.aes = FALSE, fill = cols[3], alpha=0.3) +
  geom_ribbon(data = pred4, aes(x = date, ymin = lwr, ymax = upr),
              inherit.aes = FALSE, fill = cols[4], alpha=0.3) +
  coord_cartesian(ylim = c(0, max(ylim))) +
  labs(x = "", y = "Recovered", color = "Model") +
  scale_y_continuous(breaks = seq(0, max(ylim), by = 5000),
                     minor_breaks = seq(0, max(ylim), by = 1000)) +
  scale_x_date(date_breaks = "2 day", date_labels =  "%b%d",
               minor_breaks = "1 day") +
  scale_color_manual(values = cols) +
  theme_bw() +
  theme(legend.position = "top",
        axis.text.x = element_text(angle=60, hjust=1))

Description of evolution

Positive cases and administered swabs

df = data.frame(date = COVID19$data,
                positives = c(NA, diff(COVID19$totale_casi)),
                swabs = c(NA, diff(COVID19$tamponi)))
df$x = as.numeric(df$date) - min(as.numeric(df$date)) + 1
# df$y = df$positives/df$swabs
df$y = df$positives/c(NA, zoo::rollmean(df$swabs, 2))
df = subset(df, swabs > 50)
# DT::datatable(df[,-4], )
ggplot(df, aes(x = date)) + 
  geom_point(aes(y = swabs, color = "swabs"), pch = 19) +
  geom_line(aes(y = swabs, color = "swabs")) +
  geom_point(aes(y = positives, color = "positives"), pch = 0) +
  geom_line(aes(y = positives, color = "positives")) +
  labs(x = "", y = "Number of cases", color = "") +
  scale_x_date(date_breaks = "2 day", date_labels =  "%b%d",
               minor_breaks = "1 day") +
  scale_color_manual(values = palette()[c(2,1)]) +
  theme_bw() +
  theme(legend.position = "top",
        axis.text.x = element_text(angle=60, hjust=1))

ggplot(df, aes(x = date, y = y)) + 
  geom_smooth(method = "loess", se = TRUE, col = "black") +
  geom_point(col=palette()[4]) + 
  geom_line(size = 0.5, col=palette()[4]) +
  labs(x = "", y = "% positives among admnistered swabs (two-day rolling mean)") +
  scale_y_continuous(labels = scales::percent_format(),
                     breaks = seq(0, 0.5, by = 0.05)) +
  coord_cartesian(ylim = c(0,max(df$y, na.rm = TRUE))) +
  scale_x_date(date_breaks = "2 day", date_labels =  "%b%d",
               minor_breaks = "1 day") +
  theme_bw() +
  theme(legend.position = "top",
        axis.text.x = element_text(angle=60, hjust=1))

Hospitalized and ICU patients

df = data.frame(date = COVID19$data,
                hospital = c(NA, diff(COVID19$totale_ospedalizzati)),
                icu = c(NA, diff(COVID19$terapia_intensiva)))
df$x = as.numeric(df$date) - min(as.numeric(df$date)) + 1
ggplot(df, aes(x = date, y = hospital)) + 
  geom_smooth(method = "loess", se = TRUE, col = "black") +
  geom_point(col = "orange") + 
  geom_line(size = 0.5, col = "orange") +
  labs(x = "", y = "Change hospitalized patients") +
  coord_cartesian(ylim = range(df$hospital, na.rm = TRUE)) +
  scale_y_continuous(minor_breaks = seq(min(df$hospital, na.rm = TRUE),
                                        max(df$hospital, na.rm = TRUE), 
                                        by = 100)) +
  scale_x_date(date_breaks = "2 day", date_labels =  "%b%d",
               minor_breaks = "1 day") +
  theme_bw() +
  theme(legend.position = "top",
        axis.text.x = element_text(angle=60, hjust=1))

ggplot(df, aes(x = date, y = icu)) + 
  geom_smooth(method = "loess", se = TRUE, col = "black") +
  geom_point(col = "red2") + 
  geom_line(size = 0.5, col = "red2") +
  labs(x = "", y = "Change ICU patients") +
  coord_cartesian(ylim = range(df$icu, na.rm = TRUE)) +
  scale_y_continuous(minor_breaks = seq(min(df$icu, na.rm = TRUE), 
                                        max(df$icu, na.rm = TRUE), 
                                        by = 10)) +
  scale_x_date(date_breaks = "2 day", date_labels =  "%b%d",
               minor_breaks = "1 day") +
  theme_bw() +
  theme(legend.position = "top",
        axis.text.x = element_text(angle=60, hjust=1))